Despite the prevalence of poverty in many cities, the study of extremely poor inner-city residents remains relatively neglected. This is particularly the case for African American adolescents living in extreme poverty, who arguably are at the greatest risk for negative health and life outcomes. We focus on this particularly vulnerable population, testing methodological assumptions about the Mobile Youth Survey (MYS), a multiple cohort longitudinal study of adolescents living in the Mobile, AL, Metropolitan Statistical Area. The MYS is one of the largest studies of its kind;it was begun in 1998 and represents approximately 26,700 data points collected from approximately 9,000 respondents over the past 11 years. It also uses one of the most impoverished samples reported in the literature: Median household income (1990 Census) for the neighborhoods studied was $5,190 per year, and 73% of residents in these neighborhoods lived below the poverty level. Over the years, the MYS has received funding from a variety of sources, each with a different set of goals and priorities. As a result, our sampling plan was not strictly random, but used a mixed strategy that attempted to sample as much of the population of adolescents from these neighborhoods as possible. We believe that the sample is representative, and that missing data patterns are random. However, before the MYS data can achieve its full potential, we must test these assumptions. Thus, the first two aims of this study are to assess (a) the representativeness of the MYS sample, and (b) patterns of missing data to determine whether they are Missing at Random. We are fortunate to have school records from the Mobile County Public School System and lease information from the Mobile Housing Board, which we will use to supplement the MYS data in these analyses. A third aim considers whether dropout in the MYS is higher for respondents who have engaged in sexual, drug-related, and/or violent behaviors than for respondents who have not. This is an important question that will help us not only better understand the MYS sample but also provide useful information for studying other at-risk adolescent populations. PUBLIC HEALTH RELEVANCE: The past decade has witnessed considerable progress toward many of the public health goals developed in Healthy People 2010, but others seem to defy efforts to achieve them. Tragically, many of these gains have left those living in poverty untouched, with morbidity and mortality remaining unacceptably high-particularly among inner-city youth. This project tests several methodological and substantive assumptions about the Mobile Youth Survey, an 11-year longitudinal study of very impoverished African American adolescents, to determine its suitability for studying development and risk behavior in this population.